Mechanisms of Oriented Attachment of TiO2 Nanocrystals in Vacuum

Mar 6, 2014 - ABSTRACT: Oriented attachment (OA) of nanocrystals is now widely recognized as a key process in the solution-phase growth of hierarchica...
1 downloads 0 Views 9MB Size
Download PDF
Letter pubs.acs.org/NanoLett

Mechanisms of Oriented Attachment of TiO2 Nanocrystals in Vacuum and Humid Environments: Reactive Molecular Dynamics Muralikrishna Raju,† Adri C. T. van Duin,‡ and Kristen A. Fichthorn*,†,§ †

Department of Physics, ‡Department of Mechanical and Nuclear Engineering, and §Department of Chemical Engineering, The Pennsylvania State University, University Park, Pennsylvania 16802 S Supporting Information *

ABSTRACT: Oriented attachment (OA) of nanocrystals is now widely recognized as a key process in the solution-phase growth of hierarchical nanostructures. However, the microscopic origins of OA remain unclear. We perform molecular dynamics simulations using a recently developed ReaxFF reactive force field to study the aggregation of various titanium dioxide (anatase) nanocrystals in vacuum and humid environments. In vacuum, the nanocrystals merge along their direction of approach, resulting in a polycrystalline material. By contrast, in the presence of water vapor the nanocrystals reorient themselves and aggregate via the OA mechanism to form a single or twinned crystal. They accomplish this by creating a dynamic network of hydrogen bonds between surface hydroxyls and surface oxygens of aggregating nanocrystals. We determine that OA is dominant on surfaces that have the greatest propensity to dissociate water. Our results are consistent with experiment, are likely to be general for aqueous oxide systems, and demonstrate the critical role of solvent in nanocrystal aggregation. This work opens up new possibilities for directing nanocrystal growth to fabricate nanomaterials with desired shapes and sizes. KEYWORDS: Oriented attachment, nanoparticle, titanium dioxide, anatase, water OA was first observed by Penn and Banfield16 in nanocrystalline anatase synthesized under hydrothermal conditions via the sol−gel method. In the case of hydrothermally coarsened anatase, the nanocrystals assemble into twinned or single-crystalline structures composed of several primary crystallites, suggesting a sequence of whole-particle alignment followed by interface elimination. In colloidal systems, OA could originate from intrinsic interactions between nanocrystals, such as dipole−dipole interactions, solvent-mediated interactions, such as selective adsorption and surface chemistry of liquid-phase molecules at the nanocrystal−liquid interface, or a combination of the two. By studying the aggregation in both vacuum and aqueous environments, we can isolate the roles of intrinsic forces between the nanocrystals and those mediated by the solvent and thus clarify the pathway by which OA occurs. Reactive force-field methods, such as ReaxFF, can simulate chemical reactions with higher efficiencies than ab initio MD, making them a suitable choice to model nanometer-sized particles in solvent. Here, we employ the recently developed

T

he ability to direct and control crystal growth by the oriented attachment (OA) mechanism, whereby nanocrystals interact and aggregate along specific crystallographic directions to form single- or twinned-crystal structures, is an exciting prospect that could allow the design and synthesis of hierarchical nanocrystal structures for functional devices. Crystallization by aggregation and coalescence of nanocrystal occurs frequently in natural processes as applied by organisms during biomineralization,1,2 in morphosynthesis of biomimetic inorganic materials,3 and in controlled synthesis of nanocrystals and nanowires.4−7 Remarkably, growth by OA occurs in wide variety of organic and inorganic materials exhibiting diverse particle morphologies including spheres,7,8 ellipsoids,9 cubes,4 platelets,10,11 wires,12 and hybrids,13 resulting in complex architectures ranging from quasi-one-dimensional nanowires and nanorods to hierarchical two- or three-dimensional superstructures.14,15 However, the origins and mechanisms of OA remain unresolved. Atomic-scale investigations with molecular dynamics (MD) simulations could contribute significantly to understanding growth by OA and furnish many details of this mechanism that are not accessible experimentally. © 2014 American Chemical Society

Received: December 7, 2013 Revised: March 4, 2014 Published: March 6, 2014 1836

dx.doi.org/10.1021/nl404533k | Nano Lett. 2014, 14, 1836−1842

Nano Letters

Letter

Figure 1. Snapshots of the initial approach of anatase nanocrystals in water. These snapshots were taken at (a) 187.5 ps; (b) 205 ps; (c) 212.5 ps; and (d) 225 ps.

325 Å × 125 Å in different initial configurations corresponding to various center-of-mass separations and orientations relative to one another. In this way, we allow for the possibility that many-body interactions could influence nanoparticle aggregation, although our observations indicate that OA is mediated by nanocrystal-pair interactions. Initially, the crystals were at least 30 Å apart, so they could rotate freely without touching each other. We simulated the system in the canonical (NVT) ensemble at 573 K to correspond to the temperature in experimental studies.6 The total simulation time ranged from 1.0−2.0 ns, which is long enough to observe the approach and aggregation of the nanocrystals. A typical simulation box contains ∼15 000−18 000 atoms and the simulations were run as parallel, eight-processor jobs using the ReaxFF reactive force field implementation in the ADF computational chemistry package.21 We do not observe aggregation by OA in vacuum, consistent with the results of Alimohammadi and Fichthorn.22 The aggregation of nanocrystals in Supporting Information Figure 2 is representative of the aggregation mechanism observed in vacuum. In vacuum, absence of adsorbed surface species allows the nanocrystals to aggregate along their direction of approach (Supporting Information Figure 2 and Video 1). The absence of OA in vacuum led us to hypothesize that water mediates growth by OA through adsorbed surface groups, molecularly or dissociatively adsorbed water in this case. Because MD simulations of nanocrystals at hydrothermal conditions are prohibitively slow, our aim was to achieve an aqueous environment, representative of bulk water, within the interparticle gap at close proximities. We thus investigated the amount of water needed to saturate the nanocrystal surfaces with hydroxyl groups at a temperature of 573 K to correspond to the temperature in experimental studies.6 We found that the hydroxyl coverage on the nanoparticle surfaces saturates when the total number of water molecules corresponds to 5 ML of

Ti/O/H ReaxFF parameter set, as described in Kim et al.17 and as recently applied to describe the glycine/TiO2 interface by Monti et al.18 and to various water/TiO2 interfaces by Raju et al.19 This force field was fit to a large, density functional theory (DFT) based, training set, including equations of state for bulk anatase, rutile, brookite, and higher-energy TiO2 crystals, as well as surface energies and water-dissociation energy barriers. As described in more detail in Kim et al.19 and Raju et al.,19 this force field was validated by comparisons to DFT/MD simulations and experimental results on water-dissociation reactions on various anatase and rutile surfaces of titania. This force field uses the same general and O/H ReaxFF parameters as employed in a number of previous ReaxFF descriptions,19 including proteins, alcohols and organic acids, inorganic acids, phosphates, zinc-oxides, iron-oxides, copper-oxides, aluminum, silica, and aluminosilicates, thus allowing straightforward transferability to multicomponent liquids and mixed metaloxide materials. Nanocrystalline anatase has been experimentally observed to possess a tetragonal bipyramidal morphology down to sizes of 3−5 nm.6 Indeed, density functional theory (DFT) calculations predict that anatase nanocrystals have this tetragonal bipyramidal shape, as given by the standard Wulff construction.20 We consider symmetric, charge-neutral, Wulffshaped anatase nanocrystals of two different sizes: large nanocrystals containing 2691 Ti and O atoms and small nanocrystals consisting of 840 atoms. As shown in Supporting Information Figure 1a,b, Wulff-shaped anatase crystals have two surfaces exposed to vacuum, the {101} and {001} surfaces. In addition to Wulff shapes, we consider three asymmetric nanocrystals shown in Supporting Information Figure 1c−e, which mimic possible off-Wulff shapes that could occur during crystal growth. We replicated the equilibrated nanocrystals and placed eight nanocrystals in a simulation cell with dimensions of 125 Å × 1837

dx.doi.org/10.1021/nl404533k | Nano Lett. 2014, 14, 1836−1842

Nano Letters

Letter

Figure 2. Snapshots of the alignment of anatase nanocrystals on {112} facets in water. These snapshots were taken at (a) 275 ps; (b) 300 ps; (c) 350 ps; and (d) 375 ps.

∼2.8 Å and they are separated by the first water layer adsorbed on their surfaces. When they achieve alignment, the nanocrystals remain aligned until coalescence is initiated at a contact point. To understand the origins of the alignment mechanism, we studied the interface between various aligned nanocrystals. Figure 3a shows the interface between two nanocrystals aligned on {112} facets. The area between the aligned nanocrystals is characterized by a network of hydrogen bonds between the bridging or terminal hydroxyls on one nanoparticle and the surface oxygens of the other nanoparticle. Water thus mediates the alignment of the nanocrystals along specific crystallographic directions through the formation of interparticle hydrogen bonds. This network of hydrogen bonds provides the longrange ordering necessary to bring and hold the nanocrystals in alignment and is formed when the nanocrystals align along a specific crystallographic direction. This hydrogen-bondingmediated, long-range ordering of anatase nanocrystals can be likened to the structure of DNA in which two strands of DNA are held together by hydrogen bonds that occur between complementary nucleotide base pairs. This illustrates how weak bonds, such as hydrogen bonds, can play a major role in the structure and stability of natural systems. The interface is devoid of associatively adsorbed water molecules and contains only hydroxyls and as discussed later has implications on which surfaces have propensity to exhibit OA. The interface in Figure 3a consists of 30 hydrogen bonds with an average hydrogen bond length of 1.70 Å, providing a total binding energy of 3.70 eV. Figure 3c is a transmission electron microscopy (TEM) image of wet SnO2 nanocrystals,25 which have a crystal structure similar to TiO2. Here, we see that aligned SnO2 nanocrystals are separated by spacing of the order of 1−2 lattice constants, which is consistent with the observation from our simulations that aligned nanocrystals are separated by a network of interparticle hydrogen bonds. We also note that

water coverage on all the nanoparticle surfaces at 573 K (Supporting Information Figure 3). Thus, we performed MD simulations in the NVT ensemble at this coverage and a temperature of 573 K. By simulating nanocrystals in water vapor, we facilitate their fast diffusion as they approach one another and we aim to capture their (water-mediated) interactions at subnanometer separations, where just a few water layers reside in the interparticle gap. It is observed experimentally23,24 [also see below in Figure 3c] that nanoparticles associate at these close separations before aggregating, so that our simulations may describe this associated state and subsequent aggregation. We observe numerous instances of OA in simulations including water. We find that water plays multiple roles in mediating OA, which is a two-step process consisting of the initial alignment of nanocrystals, followed by their coalescence to form a single or twinned crystal. Figure 1 shows a typical initial approach of two nanocrystals in water. Here, the presence of adsorbed surface species (molecularly and dissociatively adsorbed water) prevents immediate aggregation of the nanocrystals (Figure 1 and Supporting Information Video 2). This starkly contrasts their behavior in vacuum (cf., Supporting Information Figure 2). The approaching nanocrystals, which are initially at least 30 Å apart, associate at separations of ∼5 Å. At these close separations, they interact via the first and second water layers adsorbed on their surfaces. As shown in Figures 2 and 3, we have dense water coverage in the interparticle gap at these separations. Once in close vicinity, the nanocrystals continuously diffuse and rotate relative to each other, making transient contacts at many points until they fall into alignment (Figure 2 and Supporting Information Video 3). In all the trajectories, the nanocrystals interact in close proximity for significant periods of time and explore multiple configurations before alignment occurs. When the nanocrystals align, their separation reduces to 1838

dx.doi.org/10.1021/nl404533k | Nano Lett. 2014, 14, 1836−1842

Nano Letters

Letter

Figure 3a diffuse with a relatively constant translational speed of 0.35 Å/ps. However, within center-of-mass separations of 25 Å, the translational component increases to 0.7 Å/ps as they explore different configurations and later decreases to 0.3 Å/ps after alignment (Supporting Information Figure 4). This is in agreement with experimental observations indicating that translational speeds of the nanocrystals increase by 2−3 times as they approach each other.24,26 The higher speeds in our simulations compared to those in experiment arise from the low water density in our simulations, which allows us to observe nanocrystal aggregation over the MD time scale. Figure 3b shows the density plot of hydrogen (blue) and oxygen (red) atoms in Figure 3a, which was obtained as an average over a time period of 250 ps. Away from the interparticle gap, the density of the adsorbed hydrogen atoms forms mushroomshaped regions as they explore localized areas on the particle surfaces. However, in the area between the aligned nanocrystals, the mushroom-shaped regions crisscross, indicating that the nanocrystals continuously exchange hydrogens at the particle− particle interface. The interparticle region is dynamic with hydrogen transfers between various surface groups (bridging hydroxyls, terminal hydroxyls, and bridging oxygens) and this paves way for the aggregation of nanocrystals through expulsion of interparticle hydroxyls by regenerating water. Once aligned, the nanocrystals stay in alignment and the interparticle region is static with respect to the relative orientation of the nanocrystals. The dynamics at this interface can be reduced to hydrogen transfers between the various surface groups and the surrounding water. Aggregation occurs when surface hydroxyls combine with protons and evacuate the interparticle gap as water, leaving bare TiO2 surfaces to associate. The time scale over which all the interparticle surface hydroxyls desorb as water by recombining with protons is not accessible in our MD simulations at temperatures of 573 K. Thus, we ran NVT simulations at 1100 K to accelerate the kinetics of proton transfer for aligned nanocrystals. At these elevated temperatures, we observe aggregation of the aligned nanocrystals. The aggregation initiates at a contact point and proceeds rapidly in a zipper-like fashion, closing down the interparticle gap (Figure 4 and Supporting Information Video 4). The aggregation mechanism shown in Figure 4 is representative of the aggregation mechanism observed in water. The nanocrystals in Figure 4 stay in alignment for 250 ps until aggregation is initiated at a contact point after which aggregation proceeds rapidly finishing in 40 ps. Although we consider nanocrystals in vapor, we note that surrounding water molecules can play an important role in aggregation. In previous theoretical studies,19,27 it was observed that water in higher layers can assist in the dissociation of water adsorbed on titania surfaces by mediating proton transfer. It is thus conceivable that surrounding water molecules could facilitate the proton-hydroxyl recombination reaction that proceeds prior to nanoparticle aggregation. Additionally, water surrounding the nanocrystals could stabilize ejected water molecules through hydrogen bonding. The observed aggregation mechanism can therefore be expected to take place in bulk water, as well. We observe OA on the {112}, {001}, and {101} surfaces, as is also seen experimentally6 and representative cases are shown in Figure 5. To analyze the structure of the aggregates, we employ the “local order parameter” developed by Zhou and Fichthorn.28 The local order parameter can differentiate titanium atoms in anatase, anatase {112} twin, or rutile

Figure 3. (a) Network of hydrogen bonds between anatase nanocrystals aligned on {112} facets. The hydrogen-bond network is formed between terminal hydroxyls (in which a hydroxyl from a dissociated water binds to an under-coordinated Ti atom at the surface) or bridging hydroxyls (in which a hydrogen from a dissociated water binds to a bridging, 2-fold coordinated surface oxygen) on one nanocrystal and surface oxygens of the adjacent nanocrystal. (b) Density plot of hydrogen (blue) and oxygen (red) atoms in panel a obtained as an average over a time period of 250 ps. (c) TEM image of wet SnO2 nanocrystals. For synthesis and characterization please refer to Wang et al.25

recent cryogenic TEM experiments by Yuwano et al.23 suggest that the growth of goethite nanowires in water via OA is preceded by ferrihydrite mesocrystals or aligned crystallites with water in the interparticle gap. Analysis of particle motion leading to alignment shows that the translational speed increases as the particles approach one another and decreases once they align. For example, at centerof-mass separations greater than 30 Å, the nanocrystals in 1839

dx.doi.org/10.1021/nl404533k | Nano Lett. 2014, 14, 1836−1842

Nano Letters

Letter

Figure 4. Snapshots of the OA of anatase nanocrystals on {101} facets in water. These figures were taken at (a) 737.5 ps; (b) 750.75 ps; (c) 763.75 ps; and (d) 775 ps.

Figure 5. Structural analysis of OA aggregates on (a) {112}, (b) {112}, twin (c) {001}, and (d) {101} facets of anatase nanocrystals. Only Ti atoms are shown with anatase atoms in gray and anatase {112} twins in orange.

previous studies demonstrated that the anatase {112} and {001} surfaces dissociate water to a greater extent than the {101} surface for water coverages up to 3.0 ML, which implies a higher hydroxyl coverage on the {112} and {001} facets than on {101}.19 Also, the alignment of nanocrystals involves the formation of a network of hydrogen bonds directed normal to the nanocrystal surfaces and this requires the presence of facets with adsorbed hydroxyls. As shown in Figure 3a, the interface between aligned nanocrystals contains hydroxyls and is devoid of molecularly adsorbed water molecules. The hydrogen bonds formed by molecularly adsorbed water are not directed normal to the surface, which explains the low frequency of OA on the {101} surface. Our results may explain the occurrence of OA in

environments and can be used to study the crystallinity of the resulting aggregates. The local order parameter analysis establishes that aggregates obtained from our simulations are single crystals for coalescence on the {112}, {001}, and {101} surfaces, as shown in Figure 5. In addition, we observe the formation of anatase {112} twins, which has also been observed in experimental studies by Penn and Banfield.6 In 24 aggregation events, we observed OA 14 times on anatase {112}, 8 times on anatase {001}, and 2 times on anatase {101}. Experimentally OA has been observed to occur most commonly on {112}, occasionally on {001}, and rarely on {101}.6 We find that OA most frequently involves the surfaces that have the highest propensity to dissociate water. Our 1840

dx.doi.org/10.1021/nl404533k | Nano Lett. 2014, 14, 1836−1842

Nano Letters

Letter

The OA pathway observed in our simulations is consistent with the pathway observed in TEM studies of OA for iron oxyhydroxide nanocrystals by Li et al.24 TEM images from that study show a sequence of whole-particle alignment followed by coalescence initiated at a contact point. The iron oxyhydroxide nanocrystals continuously rotate and interact until they find a perfect lattice match. Thereafter, aggregation proceeds rapidly after a sudden jump to contact. Thus, the OA pathway observed in our simulations is likely generic to aqueous oxide materials and indicates the important role played by water and hydrogen bonding in mediating crystal growth by OA.

oxides that are known to dissociate water to a significant extent, such as SnO229,30 and ZnO.5,31 SnO2 and ZnO surfaces ionize to a greater extent than TiO2 in response to bulk solution pH32 and this may affect their aggregation mechanism. OA growth of SnO225 shows a sequence of particle alignment followed by interface elimination. Figure 3c shows aligned SnO2 nanocrystals associating at very close separations of 2−3 Å without aggregating. At these close separations, only a single layer of water could be present on the nanoparticle surfaces and can be conceived to consist only of adsorbed hydroxyls or molecular water. The water layer can facilitate the creation of hydrogen bonds which in turn bridges the adjacent nanoparticles, yielding the OA growth mechanism upon elimination of adsorbed species. This suggests a similar mechanism of alignment followed by coalescence in SnO2. This mechanism is also observed in iron oxyhydroxide nanocrystals.23,24 The anatase nanocrystal surfaces are susceptible to protonation whereby the surfaces adsorb or desorb protons in response to the bulk solution pH. The room temperature pHpzc (point of zero charge) of nanocrystalline anatase is reported33 to range from pH 5 to 7. We calculated the surface charge density using the charge description scheme in ReaxFF for the anatase nanoparticle shown in Supporting Information Figure 1d, which was equilibrated in 5 ML of water for 75 ps at 573 K. We find a negative value (∼−0.022 C/m2), suggesting that the force field is able to reproduce deprotonation at pH > pHpzc. The observed surface charge density in our simulations is reasonable, since experiments report the surface charge density at 298 K and pH 7 of anatase34,35 as ∼−0.01 C/m2 and of rutile36 as ∼−0.02 C/m2. Comparing the surface charge density predicted by the force field with experiments is nontrivial because it is dependent on temperature, the size and facets of the nanoparticle, the density of water in the simulations, and the electrolyte.34 Previous MD studies32,37 investigated the effect of temperature and pH on surface protonation and solvation structure at the rutile(110) surface and showed that surface protonation can modify the hydrogen-bonding network and thus the interfacial water structure. On silica nanoparticle surfaces, the degree of ionization has been shown to modify the interfacial water structure by changing the fraction of ions adsorbed on the surfaces.38 Similarly on anatase nanoparticles surface ionization may affect the penetration of protons into solution and modify the number of freely dissociated, mobile protons present near the surface. The effects of surface ionization on OA are not known and investigating the variation in surface ionization of anatase nanocrystals with pH and its effects on interfacial structures of water may provide a way to probe the effects of pH on OA. Penn and Banfield6 observe OA in titania hydrothermally coarsened in deionized (DI) water as well as in acidic conditions. The frequency of observed attachments in the most acidic solutions (pH 1−2) as well as in solutions adjusted using NaOH (pH 8−11) is similar to that observed in DI water. However at a pH around 3, an increased frequency of OA is observed. Our simulations are representative of the aggregation mechanism observed in DI water and the effect of buffers and pH of solution on OA may be addressable in future studies. Also of importance is the role of structure directing agents (SDA). These agents ranging from small monomer molecules to heavy-molecular-weight polymers and organic ligands are known to modify and direct OA.39,40 Future studies could target the mechanisms of OA in such systems.

■

ASSOCIATED CONTENT

S Supporting Information *

Movies are included to illustrate the approach, alignement and aggregation of anatase nanocrystals in presence of water shown in Figures 1, 2, and 3and aggregation of annatse naocrystals in vacuum shown in Supporting Figure 2. Supporting figures illustrate equilibrated structures of anatase nanocrystals, aggregation of anatase nanocrystals in vacuum, hydroxyl coverage over anatase nanocrystals as a function of water coverage at 573 K and translational speeds of anatase nanocrystals during alignment. A discussion on Ti/O/H force field development and validation is included. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: fi[email protected]. Phone: +1-814-863-4807. Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was sponsored by the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences and Biosciences (Raju and van Duin), and by the Materials Science Division, Grant DE-FG02-07ER46414 (Fichthorn). We gratefully acknowledge Hsiu-Wen Wang (Chemical Sciences Division, Oak Ridge National Laboratory) and Wei Wang (Environmental Sciences Division, Oak Ridge National Laboratory) for providing HRTEM images of SnO2 nanoparticles.

■

REFERENCES

(1) Politi, Y.; Arad, T.; Klein, E.; Weiner, S.; Addadi, L. Science 2004, 306 (5699), 1161−1164. (2) Killian, C. E.; Metzler, R. A.; Gong, Y. U. T.; Olson, I. C.; Aizenberg, J.; Politi, Y.; Wilt, F. H.; Scholl, A.; Young, A.; Doran, A.; et al. J. Am. Chem. Soc. 2009, 131 (51), 18404−18409. (3) Cölfen, H.; Antonietti, M. Angew. Chem., Int. Ed. 2005, 44 (35), 5576−5591. (4) Schliehe, C.; Juarez, B. H.; Pelletier, M.; Jander, S.; Greshnykh, D.; Nagel, M.; Meyer, A.; Foerster, S.; Kornowski, A.; Klinke, C.; et al. Science 2010, 329 (5991), 550−553. (5) Pacholski, C.; Kornowski, A.; Weller, H. Angew. Chem., Int. Ed. 2002, 41 (7), 1188−1191. (6) Penn, R. L.; Banfield, J. F. Geochim. Cosmochim. Acta 1999, 63 (10), 1549−1557.

1841

dx.doi.org/10.1021/nl404533k | Nano Lett. 2014, 14, 1836−1842

Nano Letters

Letter

(7) Cho, K.-S.; Talapin, D. V.; Gaschler, W.; Murray, C. B. J. Am. Chem. Soc. 2005, 127 (19), 7140−7147. (8) Yu, J. H.; Joo, J.; Park, H. M.; Baik, S.-I.; Kim, Y. W.; Kim, S. C.; Hyeon, T. J. Am. Chem. Soc. 2005, 127 (15), 5662−5670. (9) Bailey, J. K.; Brinker, C. J.; Mecartney, M. L. J. Colloid Interface Sci. 1993, 157 (1), 1−13. (10) Zhang, G.; Wang, W.; Lu, X.; Li, X. Cryst. Growth Des. 2008, 9 (1), 145−150. (11) Liu, Z.; Zhou, H.; Lim, Y. S.; Song, J.-H.; Piao, L.; Kim, S.-H. Langmuir 2012, 28 (25), 9244−9249. (12) Halder, A.; Ravishankar, N. Adv. Mater. 2007, 19 (14), 1854− 1858. (13) Zitoun, D.; Pinna, N.; Frolet, N.; Belin, C. J. Am. Chem. Soc. 2005, 127 (43), 15034−15035. (14) Heiligtag, F. J.; Rossell, M. D.; Suess, M. J.; Niederberger, M. J. Mater. Chem. 2011, 21 (42), 16893−16899. (15) Yang, H. G.; Zeng, H. C. Angew. Chem., Int. Ed. 2004, 43 (44), 5930−5933. (16) Penn, R. L.; Banfield, J. F. Science 1998, 281 (5379), 969−971. (17) Kim, S.-Y.; Kumar, N.; Persson, P.; Sofo, J.; van Duin, A. C. T.; Kubicki, J. D. Langmuir 2013, 29 (25), 7838−7846. (18) Monti, S.; van Duin, A. C. T.; Kim, S.-Y.; Barone, V. J. Phys. Chem. C 2012, 116 (8), 5141−5150. (19) Raju, M.; Kim, S.-Y.; van Duin, A. C. T.; Fichthorn, K. A. J. Phys. Chem. C 2013, 117 (20), 10558−10572. (20) Lazzeri, M.; Vittadini, A.; Selloni, A. Phys. Rev. B 2001, 63 (15), 155409. (21) te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Fonseca Guerra, C.; van Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. J. Comput. Chem. 2001, 22 (9), 931−967. (22) Alimohammadi, M.; Fichthorn, K. A. Nano Lett. 2009, 9 (12), 4198−4203. (23) Yuwono, V. M.; Burrows, N. D.; Soltis, J. A.; Penn, R. L. J. Am. Chem. Soc. 2010, 132 (7), 2163−2165. (24) Li, D.; Nielsen, M. H.; Lee, J. R. I.; Frandsen, C.; Banfield, J. F.; De Yoreo, J. J. Science 2012, 336 (6084), 1014−1018. (25) Wang, H.-W.; Wesolowski, D. J.; Proffen, T. E.; Vlcek, L.; Wang, W.; Allard, L. F.; Kolesnikov, A. I.; Feygenson, M.; Anovitz, L. M.; Paul, R. L. J. Am. Chem. Soc. 2013, 135 (18), 6885−6895. (26) Liao, H.-G.; Cui, L.; Whitelam, S.; Zheng, H. Science 2012, 336 (6084), 1011−1014. (27) Kumar, N.; Kent, P. R. C.; Bandura, A. V.; Kubicki, J. D.; Wesolowski, D. J.; Cole, D. R.; Sofo, J. O. J. Chem. Phys. 2011, 134 (4), 044706. (28) Zhou, Y.; Fichthorn, K. A. J. Phys. Chem. C 2012, 116 (14), 8314−8321. (29) Lee, E. J. H.; Ribeiro, C.; Longo, E.; Leite, E. R. J. Phys. Chem. B 2005, 109 (44), 20842−20846. (30) Bandura, A. V.; Kubicki, J. D.; Sofo, J. O. J. Phys. Chem. B 2008, 112 (37), 11616−11624. (31) Meyer, B.; Marx, D.; Dulub, O.; Diebold, U.; Kunat, M.; Langenberg, D.; Wöll, C. Angew. Chem., Int. Ed. 2004, 43 (48), 6641− 6645. (32) Machesky, M. L.; Předota, M.; Wesolowski, D. J.; Vlcek, L.; Cummings, P. T.; Rosenqvist, J. r.; Ridley, M. K.; Kubicki, J. D.; Bandura, A. V.; Kumar, N.; et al. Langmuir 2008, 24 (21), 12331− 12339. (33) Kosmulski, M. J. Colloid Interface Sci. 2006, 298 (2), 730−741. (34) Holmberg, J. P.; Ahlberg, E.; Bergenholtz, J.; Hassellöv, M.; Abbas, Z. J. Colloid Interface Sci. 2013, 407 (0), 168−176. (35) Sprycha, R. J. Colloid Interface Sci. 1984, 102 (1), 173−185. (36) Machesky, M. L.; Wesolowski, D. J.; Palmer, D. A.; Ridley, M. K. J. Colloid Interface Sci. 2001, 239 (2), 314−327. (37) Předota, M.; Machesky, M. L.; Wesolowski, D. J.; Cummings, P. T. J. Phys. Chem. C 2013, 117 (44), 22852−22866. (38) Patwardhan, S. V.; Emami, F. S.; Berry, R. J.; Jones, S. E.; Naik, R. R.; Deschaume, O.; Heinz, H.; Perry, C. C. J. Am. Chem. Soc. 2012, 134 (14), 6244−6256.

(39) Polleux, J.; Pinna, N.; Antonietti, M.; Hess, C.; Wild, U.; Schlögl, R.; Niederberger, M. Chem.Eur. J. 2005, 11 (12), 3541− 3551. (40) Zhang, Q.; Liu, S.-J.; Yu, S.-H. J. Mater. Chem. 2009, 19 (2), 191−207.

1842

dx.doi.org/10.1021/nl404533k | Nano Lett. 2014, 14, 1836−1842